Gene deletion in Mycobacterium tuberculosis (Mtb) and the related M. tuberculosis complex (MtbC) is emerging as a significant determinant of biological diversity. The identification of novel deletions in Mtb strains is important to serve as markers for epidemiological purposes as well as to assign roles to specific genes. As an extension of previous work, MtbC isolates from patients with tuberculosis collected during 2002/2003 from Rio de Janeiro, Brazil, were genotypically differentiated to the MtbC subspecies level. All MtbC members were typed as Mtb sensu stricto. Importantly, 30% of the Mtb isolates (n=382) failed to amplify from the IS1561' locus. PCR amplification was successful in bridging the apparent deletion. Sequence analysis of the deletion junction showed that all of the isolates tested have an identical chimeric gene which resulted from reciprocal recombination between homologous genes and an unprecedented deletion of >26 kb. Among the 9 genes deleted were two PPE genes, proteins believed to have a role in mycobacterial survival inside the immune response-derived granuloma and which confer antigenic variability. This novel Mtb deletion is designated as RD Rio. All Mtb isolates were further genotyped into principal genetic groups by katG and gyrA sequence polymorphisms. Analysis of the wild-type Mtb isolates (no RD Rio deletion) showed that isolates distributed in all three principal genetic groups. In contrast, RD Rio Mtb isolates were all typed as Group 2. When patients were segregated into those infected by RD mx or wild-type Mtb, trends emerged. Included was the finding of more rapid onset "severe" illness in patients infected with an RD Rio Mtb strain. Additionally, RD Rio Mtb is infectious and transmissible as shown by at least two instances in which epidemiological/contract and molecular genotype data strongly suggested linked-transmission. We hypothesize that RD Rio Mtb may be an emerging epidemic strain causing tuberculosis in Brazil that may have already migrated to the USA. In addition, we hypothesize that the loss of the RD Rio locus, containing the PPE genes and another five (5) coding genes, confers adaptive pathologic advantage for human disease through enhanced intracellular survival and specifically by modulation of the host immune response. This proposal seeks to further characterize the epidemiology and disease in patients with RD Rio Mtb strains, to determine spoligotype and IS6110 cluster profiles, and begin to evaluate the molecular and cellular bases underlying why this strain may have an adaptive advantages. The proposed and planned future studies may further the control of an unrecognized emerging epidemic and identify both host and pathogen factors that mediate clinical and pathogenic variability.